who/bs/10.2138 english only expert committee on …

WHO/BS/10.2138

ENGLISH ONLY

EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION

Geneva, 18 to 22 October 2010

Collaborative Study to Evaluate the Proposed 1st WHO International

Standard for Human Cytomegalovirus (HCMV) for Nucleic Acid

Amplification (NAT)-Based Assays

Jacqueline F. Fryer

1,3, Alan B. Heath

2, Rob Anderson

1, Philip D. Minor

1 and the

Collaborative Study Group *

1 Division of Virology and

2 Biostatistics

National Institute for Biological Standards and Control,

South Mimms, Potters Bar, Herts, EN6 3QG, UK

3 Study Coordinator; Tel +44 1707 641000, Fax +44 1707 641050,

E-mail [email protected]

* See Appendix 1

© World Health Organization 2010

All rights reserved. Publications of the World Health Organization can be obtained from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791 3264; fax: +41 22 791 4857; e-mail: [email protected]). Requests for permission to reproduce or translate WHO publications – whether for sale or for noncommercial distribution – should be addressed to WHO Press, at the above address (fax: +41 22 791 4806; e-mail: [email protected]).

The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for which there may not yet be full agreement. The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.

All reasonable precautions have been taken by the World Health Organization to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization be liable for damages arising from its use. The named authors alone are responsible for the views expressed in this publication.

WHO/BS/10.2138

Summary This report describes the development and worldwide collaborative study evaluation of the

candidate 1st WHO International Standard for human cytomegalovirus (HCMV) for use in the

standardisation of nucleic acid amplification techniques (NAT). Proposals for the formulation of

the candidate standard were discussed at the Standardisation of Genome Amplification

Techniques (SoGAT) Clinical Diagnostics meeting at NIBSC in June 2008. The candidate is a

whole virus preparation of the HCMV Merlin strain, formulated in a universal buffer comprising

Tris-HCl and human serum albumin, and freeze-dried for long-term stability. Thirty-two

laboratories from 14 countries participated in a collaborative study to evaluate the fitness for

purpose and potency of the candidate standard using their routine NAT-based assays for HCMV.

The freeze-dried candidate standard (Sample 1) was evaluated alongside the liquid bulk of the

candidate preparation (Sample 2), a whole virus HCMV AD169 preparation (Sample 3) and

purified Merlin DNA cloned into a bacterial artificial chromosome (Sample 4). The majority of

data sets returned were from laboratory-developed quantitative assays based on real-time PCR

technology. However, a wide range of extraction and amplification methodologies were used.

The overall mean potency estimate for the candidate standard sample 1, across the different

laboratory assays, was 5×106 (6.7 log10) 'copies/mL'. The variability in individual mean estimates

for whole virus samples 1-3 was 2 log10 (100-fold), however, the variability for the purified

DNA sample 4 was higher (>3 log10). The agreement between laboratories was markedly

improved when the potencies of the virus samples 2 and 3 were expressed relative to the

candidate standard (sample 1). In contrast, the agreement between laboratories for the purified

DNA sample 4 was not improved. This suggests that purified DNA that is not extracted

alongside the clinical samples is not suitable for standardising these types of assays. The overall

data returned from each laboratory indicates that there was no significant loss in potency upon

freeze-drying. In addition, the results obtained from accelerated thermal degradation studies at

four and eight months indicate that the candidate is extremely stable and suitable for long-term

use.

The results of the study indicate the suitability of the candidate HCMV Merlin standard as the

proposed 1st WHO International Standard for HCMV. It is therefore proposed that the candidate

standard (NIBSC code 09/162) be established as the 1st WHO International Standard for HCMV

with an assigned potency of 5×106 International Units (IU) when reconstituted in 1 mL of

nuclease-free water.

Introduction HCMV is a ubiquitous herpesvirus with a high seroprevalence worldwide. It causes disease in

the immunologically-naïve, such as newborns and infants, and immunosuppressed individuals,

particularly transplant recipients and AIDS patients. Severe and life-threatening HCMV

infections in immunocompromised individuals are managed through the administration of anti-

herpetic agents, however, all are associated with toxicity with prolonged use.

The clinical utility of viral load measurements in the diagnosis and antiviral management of

HCMV in transplant recipients has been well documented 1,2

. Two therapeutic approaches have

evolved; prophylaxis, whereby antiviral drugs are administered for a fixed period from the time

of transplant, and pre-emptive treatment, which is administered in response to an increased risk

of CMV disease. The pre-emptive approach requires diagnosis of HCMV replication, and

initiation of antiviral therapy when a predetermined level of virus in peripheral blood is reached,

prior to the appearance of clinical symptoms. Subsequently, the levels of virus are frequently

measured in order to monitor the response to and determine the duration of treatment. Although

there is no consensus on the optimal sample type or frequency of testing, both plasma and whole

blood provide prognostic information.

WHO/BS/10.2138

Consensus guidelines for the management of HCMV infection and disease in transplant

recipients have been published 1,3

. These recommend the use of NAT-based approaches in order

to determine viral load measurements in pre-emptive programmes for disease prevention. These

NAT assays measure the quantity of HCMV DNA present in a clinical sample, following

extraction of viral nucleic acid. The application and range of NAT assays used in the diagnosis

and management of HCMV varies significantly. Currently, many sites use laboratory-developed

assays based on real-time PCR technology, many of which have been described in the literature.

A range of commercial assays are also available, and comprise either analyte-specific reagents

(ASR) or assay kits specific for different amplification platforms. Each laboratory-developed or

commercial assay differs in the specimen type and nucleic acid extraction method used, as well

as in the reagents (including primers and probes) and instrumentation used for the amplification

and detection of HCMV DNA. In addition, each assay uses proprietary quantification controls to

determine the concentration of viral DNA present. These may comprise either a plasmid clone of

the PCR target, or quantified viral DNA or virus particles, and may or may not be included in the

extraction step.

Given the heterogeneity of these NAT-based assay systems, and the lack of traceability to a

standardised reference system, it is difficult to compare viral load measurements between

different laboratories and to develop uniform treatment strategies. Indeed, variability in the

performance of different assays for HCMV has been documented 4,5

. These studies have

highlighted the need for an internationally-accepted reference standard for HCMV. In 2004, the

International Herpes Management Forum called for; ‘an international quantification standard…

to compare studies using different PCR-based systems and to facilitate patient management at

multiple care centres’ 1. In the absence of such a standard, current clinical guidelines recommend

that individual laboratories establish their own viral load thresholds for HCMV management,

which are specific to their laboratory assay 1,3

. It is also recommended that the specimen type is

not changed when monitoring patients.

The World Health Organisation’s Expert Committee on Biological Standardisation establishes

reference standards for biological substances used in the prevention, treatment or diagnosis of

human disease. WHO International Standards are recognised as the highest order of reference for

biological substances, and are arbitrarily assigned a potency in International Units (IU). Their

primary purpose is to calibrate secondary references used in routine laboratory assays, in terms

of the IU, thereby providing a uniform result reporting system, and traceability of measurements,

independent of the method used 6.

Proposals for the development of the 1st WHO International Standard for HCMV were discussed

at the SoGAT Clinical Diagnostics meeting held at NIBSC in June 2008 7. Options for source

materials and formulation of the candidate standard were discussed 8. It was agreed that the

candidate standard would comprise a whole virus preparation of the prototype clinical HCMV

strain Merlin, and would be formulated in a universal buffer for further dilution in the sample

matrix appropriate to each assay. The use of whole virus would standardise the entire assay

including both extraction and DNA amplification steps. It was also agreed that the final

concentration would be in the order of 1×107 ‘copies/mL’, and would be expressed in IU when

established. The proposal was adopted into the WHO biological standardisation programme in

October 2008.

The proposed standard is intended to be used in the in vitro diagnostics field and it relates to ISO

17511:2003 Section 5.5.

WHO/BS/10.2138

Aims of study The aim of this collaborative study is to determine the potency of the candidate standard using a

range of NAT-based assays for HCMV, and to evaluate the suitability of the candidate for the

calibration of secondary reference materials and the standardisation of HCMV viral load

measurements.

Materials Candidate standard The proposed candidate standard comprises a cell-free live virus preparation of the prototype

clinical HCMV strain Merlin 10

. This low passage strain represents a well characterised, near

complete HCMV genome compared with other laboratory strains, and has been fully sequenced

(GenBank Accession number AY446894). The Merlin strain is classified as a genotype 1 virus,

based on the glycoprotein B gene UL55. Given the wide range of samples routinely tested for

HCMV, the candidate standard is formulated in a universal buffer, comprising 10 mM Tris-HCl

and human serum albumin, for further dilution in the appropriate sample matrix used in each

laboratory assay. This preparation has then been freeze-dried to ensure long-term stability.

Preparation of bulk material A tissue culture supernatant sample of HCMV Merlin strain (passage 4) was propagated in

MRC-5 cells, infecting at a multiplicity of infection of 0.1. Tissue culture fluid (passage 6) was

harvested once a cytopathic effect (CPE) was observed, and repeated until all the cells showed

CPE. The culture fluid was clarified by low speed centrifugation and virus pelleted by

ultracentrifugation. Viral pellets were pooled to make a stock of virus in 200 mL 10 mM Tris-

HCl buffer (pH 7.4), containing 0.5% human serum albumin (Tris-HSA buffer). The human

serum albumin used in the production of the candidate standard and other study samples was

derived from licensed products, and was screened and tested negative for anti-HIV-1, HBsAg,

and HCV RNA.

The concentration of the HCMV Merlin stock was determined at NIBSC, using a laboratory-

developed real-time PCR assay. Briefly, 400 µL of sample was extracted using the QIAamp®

MinElute®

Virus Spin Kit (QIAGEN, Hilden, Germany), on the QIAcube®

instrument. Five

microlitres of purified nucleic acid was then amplified by real-time PCR using the LightCycler®

480 Instrument (Roche Applied Science, Mannheim, Germany) 11

. The target was quantified

against serial dilutions of a plasmid clone of the PCR target. The HCMV DNA concentration

was also assessed at NIBSC using two commercial HCMV assays (Roche LightCycler®

CMV

Quant Kit and Nanogen Q-CMV Real Time Complete Kit), and in five clinical laboratories in

the UK using a range of laboratory-developed and commercial assays. The stock was diluted

1/8000 in Tris-HSA buffer and dispensed in 0.5 mL volumes prior to evaluation. The remainder

of the stock was stored at -80 °C until preparation of the final bulk. The geometric mean virus

concentration from all assays, in ‘copies/mL’, was used to determine a consensus HCMV

concentration for the stock.

The bulk preparation was formulated to contain approximately 1×107 HCMV 'copies/mL' in a

final volume of 6.4 L Tris-HSA buffer, and mixed for a total of 30 minutes using a magnetic

stirrer. Approximately 250 mL of the liquid bulk was dispensed in 1 mL aliquots into 2 mL

Sarstedt screw cap tubes and stored at -80 °C. The remaining bulk volume was immediately

processed for lyophilisation in order to prepare the final product, NIBSC code 09/162.

Filling and lyophilisation of candidate standard

WHO/BS/10.2138

The filling and lyophilisation of the bulk material was performed at NIBSC, and the production

summary is detailed in Table 1. The filling was performed in a Metall and Plastic

GmbH (Radolfzell, Germany) negative pressure isolator that contains the entire filling line and

is interfaced with the freeze dryer (CS150 12m2, Serail, Arguenteil, France) through a ‘pizza

door’ arrangement to maintain integrity of the operation. The bulk material was kept at 4 °C

throughout the filling process, and stirred constantly using a magnetic stirrer. The bulk was

dispensed into 5 mL screw cap glass vials in 1 ml volumes, using a Bausch & Strobel (Ilfshofen,

Germany) filling machine FVF5060. The homogeneity of the fill was determined by on-line

check-weighing of the wet weight, and vials outside the defined specification were discarded.

Filled vials were partially stoppered with halobutyl 14mm diameter cruciform closures and

lyophilised in a CS150 freeze dryer. Vials were loaded onto the shelves at -50 °C and held at this

temperature for 4 hrs. A vacuum was applied to 270 µb over 1 hr, followed by ramping to 30 µb

over 1 hr. The temperature was then raised to -40 °C, and the vacuum maintained at this

temperature for 42.5 hrs. The shelves were ramped to 25 °C over 15 hrs before releasing the

vacuum and back-filling the vials with nitrogen. The vials were then stoppered in the dryer,

removed and capped in the isolator, and the isolator decontaminated with formaldehyde before

removal of the product. The sealed vials are stored at -20 °C at NIBSC under continuous

temperature monitoring for the lifetime of the product (NIBSC to act as custodian and worldwide

distributor).

Post-fill testing Assessments of residual moisture and oxygen content, as an indicator of vial integrity after

sealing, were determined for twelve vials of freeze-dried product. Residual moisture was

determined by non-invasive near-infrared (NIR) spectroscopy (MCT 600P, Process Sensors,

Corby, UK). NIR results were then correlated to Karl Fischer (using calibration samples of the

same excipient, measured using both NIR and Karl Fischer methods) to give % w/w moisture

readings. Oxygen content was measured using a Lighthouse Infra-Red Analyser (FMS-750,

Lighthouse Instruments, Charlottesville, USA).

Samples of the liquid bulk (n=18) and freeze-dried product (n=18) were tested by HCMV real-

time PCR as described earlier, in order to determine the homogeneity of the product prior to

dispatch for collaborative study.

Stability of the freeze-dried candidate Accelerated degradation studies are underway at NIBSC in order to predict the stability of

09/162 when stored at the recommended temperature of -20 °C. Vials of freeze-dried product are

being held at -70 °C, -20 °C, +4 °C, +20 °C, +37 °C, +45 °C. At specified time points during the

life of the product, three vials will be removed from storage at each temperature and HCMV

DNA quantified by real-time PCR (as previously described). In addition, a limited assessment of

the stability of reconstituted product was performed. Reconstituted product was stored at +4 °C,

+20 °C, and +37 °C, and HCMV DNA quantified by real-time PCR after 24 and 48 hrs.

Study samples The freeze-dried candidate HCMV Merlin preparation was evaluated alongside the unprocessed

liquid bulk (used to prepare the freeze-dried candidate), a live virus preparation of the HCMV

strain AD169 12

, and a sample of purified HCMV Merlin DNA cloned into a bacterial artificial

chromosome (BAC) 13

.

The AD169 virus was propagated in MRC-5 cells as described earlier. The culture fluid was

harvested once a CPE was observed, clarified at low speed centrifugation and virus pelleted by

WHO/BS/10.2138

ultracentrifugation. Virus was then diluted to approximately 1×107 HCMV 'copies/mL' in Tris-

HSA buffer. As the prototype laboratory strain of HCMV, AD169 DNA is frequently used as a

calibrator in NAT-based assays. It has been classified as a genotype 2 virus, based on the

glycoprotein B gene.

The Merlin BAC had been prepared from the complete HCMV Merlin genome 10,13

. BAC DNA

was purified using a Nucleobond BAC100 kit (Machery-Nagel GmbH, Düren, Germany)

according to manufacturer’s instructions. The concentration of purified BAC DNA was

determined by absorbance at 260 nm, using a NanoDrop ND-1000 spectrophotometer

(NanoDrop Technologies Inc., Wilmington, DE), and diluted to 1×105 HCMV 'copies/µL' in

nuclease-free water. The purpose of including this purified HCMV DNA sample was to

investigate the effect of the extraction step on the variability in HCMV quantification.

Aliquots of AD169 (n=18) and Merlin BAC (n=18) were tested by HCMV real-time PCR (as

previously described), in order to determine the homogeneity of the samples prior to dispatch for

collaborative study. Study samples were stored at -20 °C (sample 1) and -70°C (samples 2-4)

prior to shipment to participants.

Study samples shipped to participants were coded as samples 1-4 and were as follows:

– Sample 1 - Lyophilised preparation 09/162 in a 5 mL screw cap glass vial.

– Sample 2 - 1 mL frozen liquid preparation of the HCMV Merlin bulk (used to prepare freeze-

dried candidate) in a 2 mL Sarstedt screw cap tube.

– Sample 3 - 1 mL frozen liquid whole virus preparation of HCMV AD169 in a 2 mL Sarstedt

screw cap tube.

– Sample 4 - 50 µL frozen liquid preparation of purified BAC-cloned Merlin DNA in a 0.5 mL

Sarstedt screw cap tube.

Study design The aim of the collaborative study was to evaluate the suitability and potency of the candidate

HCMV International Standard in a range of NAT based assays. Four vials each of study samples

1-4 were delivered to participating laboratories by courier on dry ice, with specific instructions

for storage and reconstitution.

Study protocol Participants were requested to test dilutions of each sample using their routine NAT-based assay

for HCMV on four separate occasions, using a fresh vial of each sample in each independent

assay. In accordance with the study protocol (Appendix 2), the lyophilized sample 1 was to be

reconstituted with 1 mL of deionised, nuclease-free molecular-grade water and left for a

minimum of 20 minutes with occasional agitation before use. Meanwhile, study samples 1-3

were to be thawed and vortexed briefly before use.

Participants were requested to dilute samples 1-3 to within the quantitative range of the assay,

using the sample matrix specific to their individual assay, and to extract each dilution prior to

amplification. Meanwhile, participants were requested to dilute sample 4 in nuclease-free water,

and add an aliquot of each dilution directly to the amplification reaction. For quantitative assays,

participants were requested to test a minimum of two serial ten-fold dilutions within the linear

range of the assay. For qualitative assays, participants were requested to test ten-fold serial

dilutions of each sample to determine the assay end-point, and then a minimum of two half-log

serial dilutions either side of the predetermined end-point, for subsequent assays.

WHO/BS/10.2138

Participants were requested to report the viral load in ‘copies/mL’ (positive/negative for

qualitative assays) for each dilution of each sample and return results including details of

methodology used to NIBSC for analysis.

Participants Study samples were sent to 32 participants representing 14 countries (Appendix 1). Participants

were selected for their experience in CMV NAT and geographic distribution. They represented

mainly clinical laboratories, but also included a range of manufacturers of in vitro diagnostic

devices (IVDs), as well as reference, research and quality assurance laboratories. All

participating laboratories are referred to by a code number, allocated at random, and not

representing the order of listing in Appendix 1. Where a laboratory returned data using different

assay methods, the results were analysed separately, as if from different laboratories, and are

referred to as, for example, laboratory 9A, 9B etc.

Statistical methods Qualitative and quantitative assay results were evaluated separately. In the case of qualitative

assays, for each laboratory and assay method, data from all assays were pooled to give a number

positive out of number tested at each dilution step. A single ‘end-point’ for each dilution series

was calculated, to give an estimate of ‘NAT detectable units/mL’, as described previously 14

. It

should be noted that these estimates are not necessarily directly equivalent to a genuine genome

equivalent number/mL. In the case of quantitative assays, analysis was based on the results

supplied by the participants. Results were reported as ‘copies/mL’ although the relationship to

genuine genome equivalence numbers is unknown. For each assay run, a single estimate of log10

‘copies/mL’ was obtained for each sample, by taking the mean of the log10 estimates of

‘copies/mL’ across replicates, after correcting for any dilution factor. A single estimate for the

laboratory and assay method was then calculated as the mean of the log10 estimates of

‘copies/mL’ across assay runs.

Overall analysis was based on the log10 estimates of ‘copies/mL’ or ‘NAT detectable units/mL’.

Overall mean estimates were calculated as the means of all individual laboratories. Variation

between laboratories (inter-laboratory) was expressed as standard deviations (SD) of the log10

estimates and % geometric coefficient of variation (%GCV) 15

of the actual estimates. Potencies

relative to sample 1, the candidate International Standard, were calculated as the difference in

estimated log10 ‘units per mL’ (test sample – candidate standard) plus a candidate assigned value

in International Units/ml (IU/mL) for the candidate standard. So for example, if in an individual

assay, the test sample is 0.5 log10 higher than the candidate standard, and the candidate standard

is assigned 6.7 log10 IU/mL, the relative potency of the test sample is 7.2 log10 IU/mL. The same

approach was used to calculate the potencies relative to sample 4, in order the evaluate the utility

of purified DNA to standardise HCMV assays.

Variation within laboratories and between assays (intra-laboratory), was expressed as standard

deviations of the log10 estimates and %GCVs of the individual assay mean estimates. These

estimates were pooled across samples 1 to 3, but calculated separately for sample 4. The

significance of the inter-laboratory variation relative to the intra-laboratory variation was

assessed by an analysis of variance.

Results and data analysis Validation of study samples and stability assessment Production data for the candidate standard sample 1 showed that the CV of the fill mass and

mean residual moisture were within acceptable limits for a WHO International Standard (Table

WHO/BS/10.2138

1). Residual oxygen content was within the NIBSC working limit of 1.1%. Evaluation of

multiple aliquots (n=18) of each study sample at NIBSC prior to dispatch indicated that the

homogeneity of HCMV content was similar for all study samples (2SD less than 0.3 log10

‘copies/mL’ for each sample).

Samples of the candidate standard 09/162 were stored at elevated temperatures, and assayed at

NIBSC concurrently with samples stored at -20 °C and -70 °C, after 4 months or 8 months

storage, by HCMV real-time PCR (as described earlier). At each time point, three vials of

samples stored at each temperature were extracted and amplified in triplicate. The mean

estimated log10 ‘copies/ml’ and differences (log10 ‘copies/ml’) from the -70 °C baseline sample

are shown in Table 2. A negative value indicates a drop in potency relative to the -70 °C baseline.

95% confidence intervals for the differences are ±0.16 log10 based on a pooled estimate of the

standard deviation between individual vial test results. Considered individually, only the

difference of +0.204 for the 45 °C samples stored for 8 months is therefore statistically

significant. However, there does appear to be a pattern of apparent increase in potency with

increasing temperature and length of storage. The reason for this is not clear. As there is no

observed drop in potency it is not possible to fit the usual Arrhenius model for accelerated

degradation studies, or obtain any predictions for the expected loss per year with long term

storage at -20 °C. However, using the ‘rule of thumb’ that the decay rate will approximately

double with every 10 °K increase in temperature (personal communication: Dr P K Philips), and

noting that there is no detectable drop in potency after 8 months at +20 °C, then there should be

no detectable difference after 64 months at -20 °C. A similar argument applied to the +37 °C

data would imply no detectable loss after 256 months (over 20 years) at -20 °C. However, with

the unexplained trend for an apparent increase in potency at the higher temperatures,

extrapolations based on the +37 °C data may not be reliable. In summary, there is no evidence of

any degradation at any temperature after storage for 8 months. It is not possible to obtain precise

estimates of any degradation rates for long term storage at -20 °C. All available data indicates

adequate stability. Subsequent testing will take place at 12 and 18 months, then at 2, 3, 4, and 5

years.

The limited assessment of the stability of reconstituted product stored at +4 °C, +20 °C, and +37

°C for 24 and 48 hrs showed that there was no marked decrease in HCMV DNA concentration in

vials stored at +20 °C and +37 °C compared with those stored at 4 °C, as determined by real-

time PCR (data not shown).

Data received

Data were received from all 32 participating laboratories. Participants performed a variety of

different assay methods, with some laboratories performing more than one assay method. In total,

data sets were received from 53 quantitative assays, and 5 qualitative assays. Apart from the

cases noted below, there were no exclusions of data.

Qualitative Assays:

Laboratories 24 and 25 used 1-log dilution steps for all 4 assays. For laboratory 24, the majority

of the results for sample 4 were positive. Estimates for this laboratory will therefore be less

precise than from those using half-log dilution steps.

Laboratory 31 had anomalous results for sample 1 in assay 4 (negative at 10-4.5

to 10-6

but

positive at 10-6.5

dilutions). These results were excluded for this assay.

Quantitative Assays:

WHO/BS/10.2138

Laboratories 2B, 4, 19B, 19C and 25 did not return results for sample 4. This was principally

because it was not possible to determine viral load without extracting the sample.

Laboratory 12A reported problems with their second assay for most replicates of samples 1, 2 &

3. This assay was excluded from further analysis.

Laboratory 16 only provided data from 2 assays. The second assay was on freeze/thawed

extractions and was excluded. The first assay did not have valid results for sample 3 (noted by

participant as possible technical error).

Laboratory 20A reported that “Samples were frozen between dilution/extraction and PCR assay”.

Laboratory 22B returned data from 4 assays, but the last 2 were after freeze-thaw cycles and

were excluded from further analysis.

For some laboratories and assays, results from individual dilutions were excluded when they

were noted as being outside the linear range of the assays.

Summary of assay methodologies The majority of participants prepared dilutions of study samples 1-3 using either plasma or

whole blood, however, urine, PBS, and nuclease-free water were also used. The extent of the

dilutions performed varied slightly between each laboratory. Extractions were predominantly

automated, and employed a range of instruments including; Abbott m2000sp, QIAGEN’s

QIAsymphony SP and RG Q, BioRobot, MDx, and EZ1, bioMérieux NucliSENS®

easyMag®

,

Roche MagNA Pure LC and COBAS®

AmpliPrep, and Siemens VERSANT®

kPCR. Manual

extraction protocols included Roche High Pure Viral Nucleic Acid Kit, Nanogen EXTRAgen®

,

QIAGEN QIAamp (Blood DNA, DNA and Viral RNA) Mini Kits, QIAGEN QIAamp DSP

Virus Kit, Cepheid affigene®

DNA Extraction Kit, and phenol-chloroform extraction.

The majority of datasets reported the use of real-time PCR technology. Seventeen participants

used commercial assays and reagents (37 data sets), while 13 participants used laboratory-

developed assays (17 data sets). Two participants used both commercial and laboratory-

developed assays (4 data sets). Commercial assays and reagents included; Roche COBAS®

AMPLICOR CMV MONITOR Test, Nanogen Q-CMV Real Time Complete Kit, Argene CMV

R-gene™ and CMV HHV6,7,8 R-gene™, QIAGEN artus CMV (LC and RG) PCR Kits, Roche

COBAS®

TaqMan®

CMV Test, Cepheid’s affigene®

CMV trender and SmartCMV™, Abbott

RealTime CMV (in development), ‘ELITech/Epoch CMV 3.0’, and Quantification of CMV

PrimerDesign™ Ltd. The range of HCMV genes targeted included; UL122/UL123 (MIE/IE19),

UL54 (DNA polymerase), UL83 (pp65), UL55 (glycoprotein B), US8, HXFL4, and UL34 and

UL80.5. Amplification platforms included; Roche LightCycler®

1.5, 2.0 and 480 systems,

COBAS®

TaqMan®

and COBAS®

AMPLICOR Analyzer, Applied Biosystems™ 7300, 7500,

7500 Fast, and 7900 HT Fast Real-Time PCR Systems, Agilent Mx3000P®

qPCR System,

QIAGEN Rotor-Gene Q, Rotor-Gene 3000 and 6000 instruments, Cepheid SmartCycler™ II and

Bio-Rad MyCycler™. Given the range of assay combinations and variables, and the fact that no

two assays were alike (apart from two laboratories using the Roche COBAS®

AMPLICOR CMV

MONITOR Test), it was not possible to group methods and perform analysis according to the

method used.

Estimated potencies of study samples The laboratory mean estimates for each study sample for quantitative assays (in log10

‘copies/mL’) and qualitative assays (in log10 ‘NAT detectable units/mL’) are shown in Tables 3

and 4 respectively. The individual laboratory mean estimates for each assay and study sample are

also shown in histogram form in Figures 1a-d. Each box represents the mean estimate from one

laboratory, and the boxes are labelled with the laboratory code number. The results from the

qualitative assays are shaded in grey.

WHO/BS/10.2138

Results for samples 1-3 show considerable variation in viral load reported between different

assays, with estimates differing by up to 2 log10 (100-fold) (Table 5). The estimates from the

qualitative assays were typically lower than those for quantitative assays. Meanwhile, the

variability for sample 4 was greater than that of samples 1-3, although this was principally due to

outlying results from five different assays (Figure 1d). Evaluation of the spread of results based

on individual assay parameters, such as the dilution matrix, showed that there was no observed

relationship between these factors and the HCMV concentration for each sample (data not

shown).

Table 5 shows the overall mean estimates for each study sample, for quantitative and qualitative

assays, along with the standard deviation (of log10 estimates) and the %GCV (of actual estimates).

For samples 1 – 3, the standards deviation for quantitative assays is approximately 0.5 log, and

%GCV is approximately 200%. These figures are consistent with the observed 2-log range of

estimates. The spread for the qualitative assays is similar. The SD and %GCV for sample 4, are

higher than those for samples 1-3, again most likely due to the outlying results.

Comparison of overall mean estimates for freeze-dried candidate sample 1 and liquid bulk

sample 2 indicates that there was no significant loss in potency upon freeze-drying (Table 5). In

addition, comparison of overall mean estimates for Merlin sample 2 and AD169 sample 3

indicates the suitability of all assays to equally quantify these two strains.

Potencies relative to sample 1 The expression of potency of samples 2-4 relative to sample 1 (as described in the statistical

methods section), allows an assessment of the suitability of the candidate standard for the

standardisation of CMV NAT assays. The relative potencies of samples 2-4 against sample 1, for

each quantitative and qualitative assay, are shown in Tables 6 and 7 respectively. Units are

expressed as candidate log10 IU/ml in both cases. The relative potencies are also shown in

histogram form in Figures 2a-c. Figures 2a and 2b show that when the mean estimates of

samples 2 and 3 are expressed relative to sample 1, there is a marked improvement in agreement

between laboratories, compared with Figures 1b and 1c. While the results from the qualitative

assays remain more variable, they are now centred around the overall mean. However, when the

mean estimates of sample 4 are expressed relative to sample 1 (Figure 2c), there is no significant

improvement in agreement between laboratories, compared with Figure 1d.

Table 8 shows the overall mean relative potency estimates (in ‘candidate log10 IU/mL’) for

samples 2-4, for quantitative and qualitative assays, along with the standard deviation (of log10

estimates) and the %GCV (of actual estimates). For the quantitative assays, the SD has reduced

from approximately 0.5 log10 to 0.12 and 0.19 log10 for samples 2 and 3 respectively. This

demonstrates that the use of a sample 1 as a standard would lead to significant reductions in

inter-laboratory variability in the estimation of CMV concentrations for clinical samples similar

to virus samples 2 and 3. Meanwhile for sample 4, the there is no reduction in the SD for

quantitative assays. As sample 1 requires extraction, and sample 4 does not, differences in

extraction efficiency between laboratories and methods will still contribute to the observed

variation between laboratories for sample 4.

Potencies relative to sample 4 The estimated concentrations of samples 1-3 were also expressed in ‘candidate IU’, relative to

sample 4, using a hypothetical unitage of 107 IU/ml for sample 4. The relative potencies of

samples 1-3 against sample 4, for each quantitative and qualitative assay, are shown in Tables 9

and 10 respectively. The results are also shown in histogram form in Figures 3a-c. These results

show that when the purified DNA sample 4 is used as a standard there is no improvement in

agreement between laboratories, as compared with Figure 1a-c. From Table 11, it can be seen

WHO/BS/10.2138

that the standard deviation between laboratories has in fact increased from around 0.5 log10 to

0.64 log10, while the %GCVs have increased to over 300%. These results show that as the

purified DNA sample 4 is not extracted alongside clinical samples, it cannot control for

differences in extraction methods or efficiency between laboratories.

Intra-laboratory variation Table 12 shows the intra-laboratory standard deviations and %GCVs for each laboratory,

calculated by pooling estimates for samples 1-3, but separately for sample 4. For all samples, the

inter-laboratory variation was greater than the intra-laboratory variation (p<0.0001). For samples

1-3 there were differences between the repeatability of laboratory estimates across assays, with

the average standard deviation being 0.11 log10 or a %GCV of 30%. For sample 4, there was a

greater range of values between laboratories, with the average standard deviation being 0.21

log10 or a %GCV of 63%.

Discussion and conclusions In this study, a range of NAT-based assays for HCMV have been used to evaluate the suitability

and potency of the candidate standard preparation as the 1st WHO International Standard for

HCMV.

The candidate standard comprises a whole virus preparation of the prototype clinical HCMV

strain Merlin. This strain was chosen as it is well characterised and more likely to represent a

clinical virus than other laboratory-adapted strains. The use of a whole virus preparation, allows

the candidate standard to be extracted alongside clinical samples thereby standardising the entire

HCMV assay. In addition, because of the range of patient samples routinely tested for HCMV,

the candidate has been formulated in a universal buffer for further dilution in the sample matrix

appropriate to each HCMV test. This allows the matrix of the candidate standard to be as similar

as possible to that of the test analyte.

In the collaborative study, the freeze-dried candidate standard was evaluated alongside whole

virus preparations of Merlin and AD169 strains. In the analysis of results, these samples

represent the samples to be calibrated by the candidate standard and enable the study to evaluate

the suitability of the candidate standard to harmonise assay results for these samples. In addition,

a purified DNA sample comprising the entire Merlin sequence cloned into a BAC was included

in the study. The purpose of including this sample was to evaluate the effect of the extraction

step on the variability of HCMV viral load measurements. It also enabled the investigation of the

utility of a purified DNA sample to standardise assays of whole virus samples.

The study results showed that all assays detected both Merlin and AD169 strains, demonstrating

the suitability of the Merlin strain for use as the candidate International Standard, and confirming

its ability to calibrate secondary references comprising the AD169 strain.

The overall mean estimate for the candidate standard sample 1 was 5×106 (6.7 log10) 'copies/mL'.

Individual laboratory mean estimates ranged from 5.4 to 7.5 log10 ‘copies/mL’. The target

concentration for the candidate standard was 1×107 ‘copies/mL’, based on preliminary testing of

the Merlin stock at NIBSC and in a selection of UK clinical laboratories. The overall mean

estimate for the liquid bulk sample 2 was similar to that of the freeze-dried sample 1, indicating

that there was no significant loss in potency upon freeze-drying. The small difference between

the overall mean estimate and the target concentration for the candidate standard is likely to be a

result of the selection of a small subset of laboratories for preliminary testing, and the large inter-

laboratory variation observed in assay results.

WHO/BS/10.2138

The overall range in laboratory mean estimates for the whole virus study samples 1-3 was 2 log10.

This variability reflects the range and differences in diagnostic testing procedures between

laboratories and is similar to levels previously reported for HCMV NAT assays 4,5

. The overall

range in laboratory mean estimates for the purified DNA sample 4 was higher. This was

unexpected since the purified DNA sample 4 was not extracted, however, the spread was

principally due to outlying results from five assays (there was no observed relationship between

these five assays). Inter-laboratory variability was significantly greater than intra-laboratory

variability. This was also reported by Pang et al., 4.

The agreement between laboratories for virus samples 2 and 3 was markedly improved when the

potencies of these study samples was expressed relative to the candidate standard (sample 1),

demonstrating the suitability of the candidate to standardise assays of whole virus samples.

However, when the purified DNA sample 4 was used as the standard, it did not lead to any

improvement in agreement between laboratories, and such a preparation would not be suitable

for the standardisation of assays of whole virus samples.

The matter of commutability of the candidate standard for clinical HCMV samples has not been

specifically assessed in this study. Commutability is affected by a range of factors including

matrix and molecular variants of the analyte (in this case HCMV DNA). The idea behind

preparing the candidate in a universal matrix for subsequent dilution in the sample matrix

appropriate to each assay was to control for matrix effects. In this study, the marked

improvement of all assay results for samples 2 and 3 when expressed relative to sample 1,

independent of the sample matrix used to dilute the samples, might suggest that this approach

does control for different sample matrices. However, it is difficult to control for differences in

HCMV forms that are present in different clinical samples. In this study, the candidate standard

is derived from a crude cell-free preparation of HCMV from cell culture, which comprises both

whole virus and naked HCMV DNA (as determined by DNase digestion experiments – data not

shown). However, patient samples derived from peripheral blood are likely to comprise a range

of HCMV forms including whole and disrupted virions, and fragmented genomic DNA, with

different forms predominating in different blood compartments. Plasma and serum samples from

renal transplant recipients have been reported to contain highly fragmented HCMV DNA 16

.

Meanwhile, whole blood samples from the same patients comprised a mixture of highly

fragmented and large DNA forms, some of which may have been derived from whole

virus. Therefore, further studies are needed to assess the commutability of this material with

respect to HCMV DNA forms present in different clinical samples.

The results obtained from accelerated thermal degradation studies at four and eight months

indicate that the candidate is stable and suitable for long-term use.

The results of this study have demonstrated that the candidate standard, NIBSC code 09/162, has

been shown to be suitable for use as a standard in a range of NAT-based assays for the

quantification of HCMV DNA. As this is proposed as the 1st WHO International Standard for

HCMV the assignment of an International Unit is arbitrary. In the case of this study a value of

5×106 International Units has been chosen as this represents the consensus estimate for the

candidate across all laboratory assays. The uncertainty can be derived from the variance of the

fill and is 0.23%.

WHO/BS/10.2138

Proposal It is proposed that the candidate standard (NIBSC code 09/162) is established as the International

Standard for HCMV with an assigned potency of 5×106 International Units when reconstituted in

1 mL of nuclease-free water. The proposed standard is intended to be used by clinical

laboratories and IVD manufacturers to calibrate secondary references used in routine NAT-based

assays for HCMV. Proposed Instructions for Use (IFU) for the product are included in Appendix

3.

Comments from participants Fifteen of thirty-two participants responded to the report. There were no disagreements with the

suitability of the candidate standard (NIBSC code 09/162) to serve as the 1st WHO International

Standard for HCMV. The majority of comments suggested editorial changes and these have been

implemented where appropriate. Two participants commented on the proposal to assign the

candidate a concentration of 6.7 log10 IU/mL, despite the initial proposal to assign a

concentration of 7 log10 IU/mL.

Acknowledgements We would like to thank Professor Gavin Wilkinson and Dr Rich Stanton, Cardiff University, UK,

for the provision of materials used in the preparation of the candidate standard and collaborative

study samples. We would also like to thank the selected UK clinical laboratories (Claire

Atkinson, Virology, Royal Free Hospital; Dr Malcolm Guiver, HPA Laboratory North West; Dr

Melvyn Smith, London South Specialist Virology Centre; Dr Rory Gunson, West of Scotland

Specialist Virology Centre; Dr Duncan Clark, Virology, Barts and The London NHS Trust) for

performing preliminary testing of the HCMV Merlin stock. We are grateful to all participants of

the collaborative study.

References 1. Razonable RR, Emery VC; 11th Annual Meeting of the IHMF (International Herpes

Management Forum). Management of CMV infection and disease in transplant patients. 27-

29 February 2004. Herpes. 2004;11(3):77-86.

2. Baldanti F, Lilleri D, Gerna G. Monitoring human cytomegalovirus infection in transplant

recipients. J Clin Virol. 2008;41(3):237-41.

3. Kotton CN, Kumar D, Caliendo AM, Asberg A, Chou S, Snydman DR, Allen U, Humar A;

Transplantation Society International CMV Consensus Group. International consensus

guidelines on the management of cytomegalovirus in solid organ transplantation.

Transplantation. 2010;89(7):779-95.

4. Pang XL, Fox JD, Fenton JM, Miller GG, Caliendo AM, Preiksaitis JK; American Society

of Transplantation Infectious Diseases Community of Practice; Canadian Society of

Transplantation. Interlaboratory comparison of cytomegalovirus viral load assays. Am J

Transplant. 2009;9:258-68.

5. Wolff DJ, Heaney DL, Neuwald PD, Stellrecht KA, Press RD. Multi-Site PCR-based CMV

viral load assessment-assays demonstrate linearity and precision, but lack numeric

standardization: a report of the association for molecular pathology. J Mol Diagn.

2009 ;11:87-92.

6. Recommendations for the preparation, characterization and establishment of international

and other biological reference standards (revised 2004). WHO Technical Report Series 2007.

Geneva, Switzerland:WHO 2007; 932,73-131.

7. Fryer JF, Minor PD. Standardisation of nucleic acid amplification assays used in clinical

diagnostics: a report of the first meeting of the SoGAT Clinical Diagnostics Working Group.

WHO/BS/10.2138

J Clin Virol. 2009;44(2):103-5.

8. SoGAT Clinical Diagnostics Meeting report;

www.nibsc.ac.uk/spotlight/sogat/clinical_diagnostics/past_meetings.aspx

9. Dolan A, Cunningham C, Hector RD, Hassan-Walker AF, Lee L, Addison C, Dargan DJ,

McGeoch DJ, Gatherer D, Emery VC, Griffiths PD, Sinzger C, McSharry BP, Wilkinson

GW, Davison AJ. Genetic content of wild-type human cytomegalovirus. J Gen Virol.

2004;85:1301-12.

10. Guiver M, Fox AJ, Mutton K, Mogulkoc N, Egan J. Evaluation of CMV viral load using

TaqMan CMV quantitative PCR and comparison with CMV antigenemia in heart and lung

transplant recipients. Transplantation. 2001;71(11):1609-15.

11. Chee MS, Bankier AT, Beck S, Bohni R, Brown CM, Cerny R, Horsnell T, Hutchison CA

3rd

, Kouzarides T, Martignetti JA, et al. Analysis of the protein-coding content of the

sequence of human cytomegalovirus strain AD169. Curr Top Microbiol Immunol.

1990;154:125-69.

12. Stanton RJ, Baluchova K, Dargan DJ, Cunningham C, Sheehy O, Seirafian S, McSharry BP,

Neale ML, Davies J, Tomasec P, Davison AJ, Wilkinson GWG. Reconstruction of the

complete human cytomegalovirus genome in a BAC reveals RL13 to be a potent inhibitor of

replication. J Clin Invest. 2010. In press.

13. Saldanha J; Lelie N and Heath A B. Establishment of the First International Standard for

Nucleic Acid Amplification Technology (NAT) assays for HCV RNA. 1999. Vox Sanquinis.

76: 149-158.

14. Kirkwood TBL. Geometric means and measures of dispersion. Biometrics 1979;35:908-9.

15. Boom R, Sol CJ, Schuurman T, Van Breda A, Weel JF, Beld M, Ten Berge IJ, Wertheim-

Van Dillen PM, De Jong MD. Human cytomegalovirus DNA in plasma and serum

specimens of renal transplant recipients is highly fragmented. J Clin Microbiol.

2002;40(11):4105-13.

WHO/BS/10.2138

Table 1. Production summary for the candidate standard (sample 1).

NIBSC code 09/162

Product name Human Cytomegalovirus

Dates of processing Filling; 25 September 2009

Lyophilisation; 25-29 September 2009

Sealing; 29 September 2009

Presentation Freeze-dried preparation in 5mL screw-cap

glass vial

Appearance Robust opaque white cake

No. of vials filled 6029

Mean fill weight (g) 1.00 (n=126)

CV of fill weight (%) 0.23

Mean residual moisture (%) 0.6 Karl Fischer, 0.41 NIR units (n=12)

CV of residual moisture (%) 7.2

Mean oxygen content (%) 0.22 (n=12)

CV of oxygen content (%) 40.6

No. of vials available to

WHO

5100

Table 2. Stability of 09/162 at 4 and 8 months.

Mean log10 ‘copies/mL’ Difference in log10 ‘copies/mL’

from -70°C baseline sample

Temperature

(°C)

4 months 8 months 4 months 8 months

-70 6.92 6.77 - -

-20 6.92 6.78 0.008 0.015

+4 6.86 6.72 -0.054 -0.048

+20 6.96 6.84 0.044 0.071

+37 7.02 6.91 0.103 0.141

+45 7.07 6.97 0.155 0.204

WHO/BS/10.2138

Table 3. Laboratory Mean Estimates from Quantitative Assays (log10 ‘copies/mL’).

Sample Participant

S1 S2 S3 S4 01 6.91 6.92 6.93 7.26

02A 6.05 6.14 5.96 7.81

02B 6.93 7.00 6.99 .

03 6.25 6.30 6.12 7.20

04 7.24 7.14 7.09 .

05 5.71 5.89 5.99 7.21

06 6.31 6.24 6.16 6.80

07 6.03 5.96 6.00 6.56

08 6.74 6.90 6.84 8.81

09A 6.78 6.85 6.68 7.24

09B 7.24 7.20 7.00 7.34

09C 7.04 6.99 6.80 7.21

09D 6.66 6.59 6.43 6.80

09E 7.21 7.29 7.21 7.23

09F 7.32 7.46 7.39 7.15

09G 7.30 7.39 7.35 7.21

09H 7.13 7.03 7.10 6.87

09I 6.67 6.78 6.82 7.21

09J 6.84 6.89 6.85 6.73

09K 7.06 7.03 7.08 7.23

09L 7.29 7.35 7.34 7.15

09M 7.04 7.11 7.11 7.21

09N 6.92 6.74 6.83 6.87

11 7.04 7.11 7.06 7.18

12A 6.31 6.38 6.05 7.43

12B 6.36 6.35 6.57 6.55

13 5.65 5.77 5.73 5.06

14 6.76 6.78 6.76 7.34

15 6.95 6.87 6.82 6.67

16 5.97 5.58 . 6.91

17 6.66 6.75 6.73 7.64

18 6.15 5.99 5.97 5.81

19A 6.65 6.53 6.47 7.37

19B 6.49 6.42 6.36 .

19C 6.76 6.57 6.49 .

20A 6.57 6.58 7.32 7.19

20B 6.18 6.22 6.66 6.88

21 7.46 7.53 7.36 7.14

22A 6.02 5.84 6.12 6.86

22B 6.11 6.09 6.36 8.39

23 6.94 7.02 7.15 7.08

24 6.09 6.13 6.06 7.10

25 7.05 6.91 6.87 .

26A 6.82 6.75 6.65 6.41

26B 7.03 6.97 6.95 7.91

27 7.23 7.30 7.30 7.69

28A 6.85 6.65 6.66 6.55

28B 6.52 6.72 6.24 7.03

29 7.08 7.13 7.10 7.79

30A 6.67 6.82 6.72 6.09

30B 7.05 7.24 7.23 8.02

31 6.24 6.13 6.03 6.69

32 7.43 7.41 7.39 7.49

WHO/BS/10.2138

Table 4. Laboratory Mean Estimates from Qualitative Assays (log10 ‘NAT detectable

units/mL’).

Sample Laboratory

S1 S2 S3 S4

05 5.39 5.53 5.53 6.92

10 6.55 6.38 6.62 6.68

24 5.96 5.83 6.02 7.16

25 5.98 6.43 5.98 6.86

31 6.15 5.47 5.18 6.51

Table 5. Overall Mean Estimates and Inter-Laboratory Variation (log10 ‘copies/mL’ for

quantitative or ‘NAT-detectable units/mL’ for qualitative assays).

Sample Assay No.

data sets Mean SD %GCV Min Max

qualitative 5 6.01 0.42 161 5.39 6.55 S1

quantitative 53 6.71 0.46 188 5.65 7.46

qualitative 5 5.93 0.46 185 5.47 6.43 S2

quantitative 53 6.71 0.49 207 5.58 7.53

qualitative 5 5.86 0.54 249 5.18 6.62 S3

quantitative 52 6.72 0.46 190 5.73 7.39

qualitative 5 6.82 0.25 77 6.51 7.16 S4

quantitative 48 7.11 0.61 307 5.06 8.81

WHO/BS/10.2138

Table 6. Laboratory Estimates of Potency Relative to S1 from Quantitative Assays log10

IU/mL taking Sample 1 as 5×106 (6.7 log10 ) IU/mL.

Sample Participant

S2 S3 S4 01 6.70 6.71 7.05

02A 6.79 6.61 8.45

02B 6.77 6.76 .

03 6.75 6.57 7.64

04 6.60 6.55 .

05 6.88 6.98 8.20

06 6.63 6.56 7.20

07 6.63 6.67 7.23

08 6.87 6.81 8.77

09A 6.77 6.61 7.16

09B 6.66 6.46 6.81

09C 6.64 6.46 6.87

09D 6.63 6.47 6.83

09E 6.78 6.70 6.72

09F 6.84 6.78 6.54

09G 6.80 6.75 6.61

09H 6.60 6.67 6.44

09I 6.81 6.86 7.24

09J 6.75 6.70 6.59

09K 6.67 6.72 6.87

09L 6.76 6.75 6.57

09M 6.77 6.77 6.87

09N 6.51 6.61 6.65

11 6.77 6.72 6.84

12A 6.77 6.44 7.82

12B 6.69 6.91 6.89

13 6.82 6.79 6.11

14 6.72 6.70 7.29

15 6.62 6.57 6.42

16 6.31 . 7.64

17 6.79 6.77 7.68

18 6.54 6.52 6.36

19A 6.59 6.52 7.43

19B 6.63 6.56 .

19C 6.51 6.43 .

20A 6.71 7.46 7.32

20B 6.74 7.18 7.41

21 6.77 6.60 6.38

22A 6.52 6.80 7.55

22B 6.69 6.95 8.99

23 6.78 6.90 6.83

24 6.74 6.66 7.70

25 6.57 6.52 .

26A 6.63 6.53 6.29

26B 6.65 6.62 7.58

27 6.77 6.77 7.16

28A 6.50 6.50 6.39

28B 6.89 6.42 7.20

29 6.75 6.72 7.40

30A 6.86 6.75 6.13

30B 6.90 6.88 7.67

31 6.59 6.49 7.15

32 6.68 6.66 6.76

WHO/BS/10.2138

Table 7. Laboratory Estimates of Potency Relative to S1 from Qualitative Assays log10 IU/mL

taking Sample 1 as 5×106 (6.7 log10 ) IU/mL.

Sample Laboratory

S2 S3 S4

05 6.84 6.84 8.23

10 6.52 6.77 6.83

24 6.57 6.76 7.90

25 7.16 6.70 7.58

31 6.02 5.73 7.06

Table 8. Overall Mean Estimates and Inter-Laboratory Variation for Potency Relative to

Sample 1 log10 IU/mL taking Sample 1 as 5×106 (6.7 log10 ) IU/mL.

Sample Assay No.


qualitative 5 6.62 0.42 163 6.02 7.16

quantitative 53 6.70 0.12 31 6.31 6.90 S2

combined 58 6.69 0.16 44 6.02 7.16

qualitative 5 6.56 0.47 192 5.73 6.84

quantitative 52 6.69 0.19 56 6.42 7.46 S3

combined 57 6.68 0.23 68 5.73 7.46

qualitative 5 7.52 0.58 280 6.83 8.23

quantitative 48 7.12 0.64 341 6.11 8.99 S4

combined 53 7.16 0.64 340 6.11 8.99

WHO/BS/10.2138

Table 9. Laboratory Estimates of Potency Relative to S4 from Quantitative Assays log10

IU/mL taking Sample 4 as 107 (7.0 log10 ) IU/mL.

Sample Participant

S1 S2 S3 01 6.65 6.66 6.67

02A 5.25 5.34 5.15

02B

03 6.06 6.10 5.92

04

05 5.50 5.68 5.79

06 6.50 6.43 6.36

07 6.47 6.39 6.44

08 4.93 5.10 5.04

09A 6.54 6.61 6.45

09B 6.89 6.85 6.66

09C 6.83 6.78 6.59

09D 6.87 6.79 6.64

09E 6.98 7.06 6.98

09F 7.16 7.30 7.24

09G 7.09 7.18 7.14

09H 7.26 7.16 7.23

09I 6.46 6.57 6.61

09J 7.11 7.16 7.11

09K 6.83 6.80 6.85

09L 7.13 7.20 7.19

09M 6.83 6.90 6.90

09N 7.05 6.87 6.96

11 6.86 6.93 6.88

12A 5.88 5.95 5.62

12B 6.81 6.80 7.02

13 7.59 7.71 7.68

14 6.41 6.43 6.41

15 7.28 7.20 7.15

16 6.06 5.68 .

17 6.02 6.11 6.08

18 7.34 7.18 7.16

19A 6.27 6.16 6.10

19B

19C

20A 6.38 6.39 7.13

20B 6.29 6.34 6.77

21 7.32 7.39 7.22

22A 6.15 5.98 6.25

22B 4.71 4.70 4.97

23 6.87 6.94 7.07

24 6.00 6.04 5.96

25

26A 7.41 7.34 7.24

26B 6.12 6.07 6.04

27 6.54 6.61 6.60

28A 7.31 7.11 7.11

28B 6.50 6.69 6.21

29 6.30 6.35 6.31

30A 7.57 7.73 7.62

30B 6.03 6.23 6.21

31 6.55 6.45 6.34

32 6.94 6.92 6.90

WHO/BS/10.2138

Table 10. Laboratory Estimates of Potency Relative to S4 from Qualitative Assays log10

IU/mL taking Sample 4 as 107 (7.0 log10 ) IU/mL.

Sample Laboratory

S1 S2 S3

05 5.47 5.61 5.61

10 6.87 6.70 6.94

24 5.80 5.67 5.86

25 6.12 6.58 6.12

31 6.64 5.97 5.68

Table 11. Overall Mean Estimates and Inter-Laboratory Variation for Potency Relative to

Sample 4 log10 IU/mL taking Sample 4 as 107 (7.0 log10 ) IU/mL.

Sample Assay No.


qualitative 5 6.18 0.58 280 5.47 6.87

quantitative 48 6.58 0.64 341 4.71 7.59 S1

combined 53 6.54 0.64 340 4.71 7.59

qualitative 5 6.10 0.51 221 5.61 6.70

quantitative 48 6.59 0.64 333 4.70 7.73 S2

combined 53 6.54 0.64 334 4.70 7.73

qualitative 5 6.04 0.54 246 5.61 6.94

quantitative 47 6.60 0.63 325 4.97 7.68 S3

combined 52 6.54 0.64 334 4.97 7.68

WHO/BS/10.2138

Table 12. Intra-Laboratory standard deviation of log10 ‘copies/mL’ and %GCV for quantitative

assays. Samples 1-3 Sample 4 Participant

SD %GCV SD %GCV

01 0.12 33 0.13 35

02A 0.10 27 0.07 17

02B 0.09 23 - -

03 0.10 24 0.13 35

04 0.10 25 - -

05 0.13 34 0.04 10

06 0.08 21 0.31 103

07 0.12 32 0.11 30

08 0.10 27 0.09 23

09A 0.10 27 0.12 32

09B 0.06 14 0.14 38

09C 0.09 23 0.07 18

09D 0.11 30 0.10 27

09E 0.06 16 0.12 32

09F 0.09 23 0.16 44

09G 0.05 12 0.07 18

09H 0.08 19 0.11 28

09I 0.10 25 0.07 18

09J 0.13 35 0.23 68

09K 0.08 20 0.12 32

09L 0.09 22 0.16 44

09M 0.07 18 0.07 18

09N 0.12 31 0.11 28

11 0.01 3 0.09 24

12A 0.22 66 0.22 65

12B 0.03 6 0.16 44

13 0.11 28 0.32 108

14 0.03 6 0.13 35

15 0.05 12 0.05 13

16 - - - -

17 0.17 48 0.08 20

18 0.05 12 0.09 24

19A 0.08 20 0.07 18

19B 0.15 42 - -

19C 0.06 15 - -

20A 0.16 45 0.30 97

20B 0.12 33 0.07 18

21 0.09 24 0.37 134

22A 0.36 132 0.06 14

22B 0.16 45 0.05 12

23 0.06 14 0.13 35

24 0.08 21 0.29 96

25 0.05 11 - -

26A 0.07 16 0.50 215

26B 0.04 9 0.72 427

27 0.10 25 0.14 38

28A 0.05 13 0.02 5

28B 0.10 25 0.10 26

29 0.07 17 0.10 25

30A 0.24 72 0.17 47

30B 0.08 21 0.09 22

31 0.14 37 0.34 117

32 0.06 14 0.47 194

Mean 0.11 30 0.21 63

WHO/BS/10.2138

Figure legends

Figure 1. Individual laboratory mean estimates for study samples 1-4 obtained using

qualitative or quantitative NAT assays. Each box represents the mean estimate from each

laboratory assay and is labelled with the laboratory code number. The results from the qualitative

assays are shaded in grey.

Figure 2. Relative potencies of samples 2-4 against sample 1, for each quantitative and

qualitative assay. Units are expressed as candidate log10 IU/mL in both cases. Each box

represents the relative potency for each laboratory assay and is labelled with the laboratory code

number. The results from the qualitative assays are shaded in grey.

Figure 3. Estimated concentrations of samples 1-3 expressed in IU, relative to sample 4, using

a hypothetical unitage of 107 IU/mL for sample 4. Each box represents the relative potency for

each laboratory assay and is labelled with the laboratory code number. The results from the

qualitative assays are shaded in grey.

WHO/BS/10.2138

Figure 1

a

Sample 1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40

NAT Detectable Units/ml or copies/ml (log10)

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

05 05

13

02A

07

16

22A

22B

24

24

25

03

06

12A

12B

18

20B

31

31

10

19B

20A

28B

08

09A

09D

09I

09J

14

17

19A

19C

26A

28A

30A

01

02B

09C

09K

09M

09N

11

15

23

25

26B

29

30B

04

09B

09E

09F

09G

09H

09L

27

21

32

WHO/BS/10.2138

b

Sample 2

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

05

16

31

13

22A

24

05

07

18

22B

02A

03

06

12B

20B

24

31

09D

10

12A

19A

19B

19C

20A

25

09A

09I

09N

14

15

17

26A

28A

28B

30A

01

02B

08

09C

09H

09J

09K

09M

11

23

25

26B

04

09B

09E

09L

27

29

30B

09F

09G

21

32

WHO/BS/10.2138

c

Sample 3

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

31 05 13 02A

03

05

07

12A

18

22A

24

24

25

31

06

19B

22B

28B

09D

10

12B

19A

19C

08

09A

09C

09I

09J

09N

14

15

17

20B

25

26A

28A

30A

01

02B

04

09B

09H

09K

09M

11

26B

29

09E

09G

09L

20A

21

23

27

30B

09F

32

WHO/BS/10.2138

d

Sample 4

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

13 18 30A 07

12B

26A

28A

31

06

09D

09H

09J

09N

10

15

22A

25

31

05

16

20B

23

24

28B

01

03

05

09A

09B

09C

09E

09F

09G

09I

09K

09L

09M

11

14

19A

20A

21

24

12A

32

02A

17

27

29

26B

30B

22B 08

WHO/BS/10.2138

Figure 2

a

Sample 2 - Relative to Sample 1

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40

Candidate IU/ml (log10)

4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5

31 16 04

09H

09N

10

15

18

19A

19C

22A

24

25

28A

31

01

02A

02B

03

05

06

07

08

09A

09B

09C

09D

09E

09F

09G

09I

09J

09K

09L

09M

11

12A

12B

13

14

17

19B

20A

20B

21

22B

23

24

26A

26B

27

29

30A

32

05

28B

30B

25

WHO/BS/10.2138

b


0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5

31 02A

03

04

06

09A

09B

09C

09D

09N

12A

15

18

19A

19B

19C

21

25

26A

26B

28A

28B

31

01

02B

05

07

08

09E

09F

09G

09H

09I

09J

09K

09L

09M

10

11

13

14

17

22A

24

24

25

27

29

30A

32

05

12B

22B

23

30B

20B 20A

WHO/BS/10.2138

c


0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5

13 18

26A

30A

09F

09G

09H

09J

09L

15

21

28A

09B

09C

09D

09E

09K

09M

09N

10

11

23

32

01

12B

31

06

07

09A

09I

14

20A

27

28B

31

19A

20B

22A

25

26B

29

03

12A

16

17

24

30B

24 05

05

02A 08 22B

WHO/BS/10.2138

Figure 3

a


0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

22B 08 02A 05

05

24 03

12A

16

17

24

25

26B

30B

19A

20B

22A

29

06

07

09A

09I

14

20A

27

28B

31

01

09C

09D

09K

09M

10

11

12B

23

31

09B

09E

09G

09J

09N

32

09F

09H

09L

15

18

21

28A

13

26A

30A

WHO/BS/10.2138

b


0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

22B 08 02A 05 05

16

24

03

12A

17

22A

24

26B

31

19A

20B

29

30B

06

07

09A

09I

14

20A

25

27

31

01

09B

09C

09D

09K

09N

10

12B

28B

09E

09M

11

23

28A

32

09F

09G

09H

09J

09L

15

18

26A

21 13

30A

WHO/BS/10.2138

c


0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

40


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0

08

22B

02A 05

12A

05

24

31

03

17

19A

24

25

26B

06

22A

28B

29

30B

31

07

09A

09C

09I

14

27

01

09B

09D

09K

20B

09E

09J

09M

09N

10

11

12B

23

28A

32

09F

09G

09H

09L

15

18

20A

21

26A

30A 13

WHO/BS/10.2138

Appendix 1

Collaborative study participants (In alphabetical order by country)

Name Laboratory Country

Prof. Dr. Harald Kessler Mol. Diagnostics Lab / IHMEM / Medical Univ. of

Graz, Graz

Austria

Prof. William Rawlinson Virology Division, SEALS Microbiology, Randwick Australia

Dr Guy Boivin Centre de recherché en infectiologie-CHUQ, Québec Canada

Dr Jutta Preiksaitis

Dr Xiao-Li Pang

Provincial Laboratory for Public Health / University of

Alberta, Alberta

Canada

Côme Barranger Argene SA, Verniolle France

Dr Sophie Alain French National Reference Centre for Cytomegalovirus,

Limoges

France

Dr Céline Bressollette-Bodin Virology Laboratory, Nantes University Hospital,

Nantes

France

Prof. Dr. Klaus Hamprecht Institute of Medical Virology, University Hospital of

Tübingen, Tübingen

Germany

Dr Thomas Grewing QIAGEN Hamburg GmbH, Hamburg Germany

Dr Pantelis Constantoulakis Locus Medicus SA, Athens Greece

Dr Valeria Ghisetti Laboratory of Microbiology and Virology, Amedeo di

Savoia Hospital,

Turin

Italy

Dr Maria R Capobianchi

Dr Isabella Abbate

Laboratory of Virology, National Institute for Infectious

Diseases "L. Spallanzani", Rome

Italy

Dr Cristina Olivo Nanogen Advanced Diagnostics, Buttigliera Alta Italy

Dr Tiziana Lazzarotto Operative Unit of Microbiology, Laboratory of

Virology, Bologna

Italy

Dr Fausto Baldanti Molecular Virology Unit, Virology and Microbiology,

Fondazione IRCCS Policlinico San Matteo, Pavia

Italy

Dr Naoki Inoue Laboratory of Herpesviruses, Department of Virology I,

National Institute of Infectious Diseases, Tokyo

Japan

Prof. Fredrik Müller Department of Microbiology, Oslo University Hospital,

Rikshospitalet,

Norway

Dr Craig Corcoran Ampath Pathology Laboratories,

Pretoria

South Africa

Dr Diana Hardie Diagnostic Virology Laboratory Groote Schuur

Hospital, Cape Town

South Africa

Dr Jacqueline Prieto Cepheid AB, Bromma Sweden

Dr Rob Schuurman

Dr Anton van Loon

University Medical Center Utrecht, Dept. Virology,

Utrecht

The Netherlands

Dr Shiaolan Ho Abbott Molecular, Inc., Des Plaines USA

Dr David Hillyard ARUP Laboratories, Inc. (University of Utah

enterprise), Salt Lake City

USA

Dr Richard Hodinka Clinical Virology Laboratory, Children’s Hospital of

Philadelphia, Philadelphia

USA

Dr Marie Louise Landry Clinical Virology Laboratory, Yale New Haven

Hospital, New Haven

USA

Dr Angela Caliendo Emory University Hospital, Altanta USA

Dr Nell Lurain NIH/DAIDS/NIAID Viral Quality Assurance

Laboratory, Rush University Medical Center, Chicago

USA

Dr Lee Sung Roche Molecular Systems Inc., Pleasanton USA

Dr Margaret Gulley University of North Carolina, Chapel Hill USA

Claire Atkinson Department of Virology, Royal Free Hospital, London UK

Dr Jon Bible GSTS Pathology, London UK

Dr Malcolm Guiver HPA Laboratory NorthWest, Department of Virology,

Manchester

UK

WHO/BS/10.2138

Appendix 2

Study protocol

WHO/BS/10.2138

WHO/BS/10.2138

Appendix 3

Proposed instructions for use

WHO/BS/10.2138

= = =

who/bs/10.2138 english only expert committee on …

Documents